update

README.md

@@ -1,9 +1,9 @@
 ---
-title: Al Tools
-emoji: 📚
+title: Test Create Docker
+emoji: 🐢
 colorFrom: red
-colorTo: red
-sdk: static
+colorTo: indigo
+sdk: docker
 pinned: false
 ---
 

active_learning.py

@@ -0,0 +1,31 @@
+
+import numpy as np
+
+def get_initial_sample(unlabeled_data, num_query):
+
+    # print(len(unlabeled_data))
+    # print(unlabeled_data)  
+
+    uncertain_samples = np.random.choice(len(unlabeled_data), size=num_query, replace=False)
+    return uncertain_samples
+
+def get_uncertain_sample(
+    labeled_data, unlabeled_data, num_query
+):
+    # print(len(labeled_data))
+    # print(labeled_data)
+
+    # print(len(unlabeled_data))
+    # print(unlabeled_data)    
+
+    uncertain_samples = np.random.choice(len(unlabeled_data), size=num_query, replace=False)
+    print(uncertain_samples)
+
+    return uncertain_samples
+
+def get_stopping_conditioon(
+    labeled_data, eval_metrics
+):
+    print(eval_metrics)
+
+    return True

app.py

@@ -0,0 +1,101 @@
+from flask import Flask, render_template, request, redirect, url_for, send_file, make_response
+import pandas as pd
+import io
+import csv
+import numpy as np
+
+from active_learning import get_initial_sample, get_uncertain_sample
+from datasets import Dataset, concatenate_datasets
+
+app = Flask(__name__)
+
+# Global variables to store data and labels
+uploaded_data = None
+labels = []
+al_step = 0
+data_amount = 100
+dataset_labeled_prev = None
+#num_query = 5
+
+@app.route("/", methods=["GET", "POST"])
+def upload():
+    global uploaded_data, labels, dataset_unlabeled, al_step, num_query
+
+    if request.method == "POST":
+        # Upload the data file
+        num_query = int(request.form.get('query_number'))
+        data_file = request.files["data_file"]
+        if data_file:
+            df = pd.read_csv(data_file)
+            unlabeled_data = Dataset.from_pandas(df)
+
+            # get_initial_sample()
+            uncertain_samples = get_initial_sample(unlabeled_data, num_query)
+
+            all_list = np.arange(0, len(unlabeled_data))
+            unlabeled_list = np.delete(all_list, uncertain_samples)
+
+            dataset_labeled = unlabeled_data[uncertain_samples]
+            dataset_unlabeled = unlabeled_data[unlabeled_list]
+
+            dataset_unlabeled =  Dataset.from_dict(dataset_unlabeled)
+    
+            uploaded_data = dataset_labeled
+
+        # Upload the labels file
+        labels_file = request.files["labels_file"]
+        if labels_file:
+            df_labels = pd.read_csv(labels_file)
+            labels = df_labels["label"].tolist()
+
+        return redirect(url_for("annotate"))
+
+    return render_template("upload.html")
+
+@app.route("/annotate", methods=["GET", "POST"])
+def annotate():
+    global uploaded_data, labels, dataset_unlabeled, al_step, dataset_labeled_prev, num_query
+
+    al_process = "Run get_initial_sample()..."  
+
+    if request.method == "POST":
+        al_step += 1
+        al_process = "Run get_uncertain_sample()..."
+        
+        annotated_data = request.form.getlist("annotations")
+        df = pd.DataFrame(uploaded_data)
+        df["label"] = annotated_data
+        dataset_labeled = Dataset.from_pandas(df)
+        
+        print(f"AL STEP #{al_step}")
+        print(dataset_labeled)
+
+        if al_step > 1:
+            dataset_labeled = concatenate_datasets([dataset_labeled_prev, dataset_labeled])
+        
+        # get_uncertain_sample()
+        uncertain_samples = get_uncertain_sample(dataset_labeled, dataset_unlabeled, num_query)
+
+        dataset_labeled_prev = dataset_labeled
+
+        unlabeled_list = np.arange(0, data_amount - al_step*num_query)
+        unlabeled_list = np.delete(unlabeled_list, uncertain_samples)
+
+        dataset_labeled_next = dataset_unlabeled[uncertain_samples]
+        dataset_unlabeled = dataset_unlabeled[unlabeled_list]
+
+        uploaded_data = dataset_labeled_next
+
+        dataset_unlabeled =  Dataset.from_dict(dataset_unlabeled)
+
+    return render_template(
+                "annotate.html", 
+                data=uploaded_data, 
+                labels=labels, 
+                al_step=al_step,
+                n_unlabeled=len(dataset_unlabeled), 
+                al_process=al_process
+            )
+
+if __name__ == "__main__":
+    app.run(host="0.0.0.0", port=7860)

index.html

@@ -1,19 +0,0 @@
-<!DOCTYPE html>
-<html>
-	<head>
-		<meta charset="utf-8" />
-		<meta name="viewport" content="width=device-width" />
-		<title>My static Space</title>
-		<link rel="stylesheet" href="style.css" />
-	</head>
-	<body>
-		<div class="card">
-			<h1>Welcome to your static Space!</h1>
-			<p>You can modify this app directly by editing <i>index.html</i> in the Files and versions tab.</p>
-			<p>
-				Also don't forget to check the
-				<a href="https://huggingface.co/docs/hub/spaces" target="_blank">Spaces documentation</a>.
-			</p>
-		</div>
-	</body>
-</html>

requirements.txt

@@ -0,0 +1,3 @@
+flask==3.0.0
+requests==2.31.*
+datasets

static/css/styles.css

@@ -0,0 +1,69 @@
+/* Style the header */
+h1 {
+    text-align: center;
+    font-size: 24px;
+    margin-bottom: 20px;
+}
+
+/* Style the form container */
+form {
+    width: 60%;
+    margin: 0 auto;
+    padding: 20px;
+    border: 1px solid #ccc;
+    border-radius: 5px;
+    background-color: #f9f9f9;
+}
+
+/* Style the file input fields */
+/* input[type="file"] {
+    display: block;
+    margin: 10px 0;
+} */
+
+
+
+/* Style the table */
+table {
+    width: 100%;
+    border-collapse: collapse;
+    margin-top: 20px;
+}
+
+/* Style table headers */
+th {
+    background-color: #333;
+    color: white;
+    text-align: left;
+}
+
+/* Style table data rows */
+tr:nth-child(even) {
+    background-color: #f2f2f2;
+}
+
+/* Style table data cells */
+td {
+    padding: 8px;
+    border: 1px solid #ddd;
+}
+
+/* Style the submit button */
+button[type="submit"] {
+    display: block;
+    margin: 20px 0;
+    padding: 10px 20px;
+    background-color: #333;
+    color: white;
+    border: none;
+    border-radius: 5px;
+    cursor: pointer;
+}
+
+/* Style the dropdown select */
+select {
+    width: 100%;
+    padding: 8px;
+    border: 1px solid #ccc;
+    border-radius: 5px;
+}

style.css

@@ -1,28 +0,0 @@
-body {
-	padding: 2rem;
-	font-family: -apple-system, BlinkMacSystemFont, "Arial", sans-serif;
-}
-
-h1 {
-	font-size: 16px;
-	margin-top: 0;
-}
-
-p {
-	color: rgb(107, 114, 128);
-	font-size: 15px;
-	margin-bottom: 10px;
-	margin-top: 5px;
-}
-
-.card {
-	max-width: 620px;
-	margin: 0 auto;
-	padding: 16px;
-	border: 1px solid lightgray;
-	border-radius: 16px;
-}
-
-.card p:last-child {
-	margin-bottom: 0;
-}

templates/annotate.html

@@ -0,0 +1,89 @@
+<!DOCTYPE html>
+<html>
+
+<head>
+    <title>Annotation Page</title>
+
+    <link rel="stylesheet" type="text/css" href="{{ url_for('static', filename='css/styles.css') }}">
+
+    <style>
+        /* Style for selected rows */
+        .selected {
+            background-color: #a0e57e !important;
+            /* Change this color as desired */
+        }
+
+        #splash-screen {
+            position: fixed;
+            top: 0;
+            left: 0;
+            width: 100%;
+            height: 100%;
+            background-color: #ffffff;
+            display: flex;
+            justify-content: center;
+            align-items: center;
+        }
+
+        /* Style for the loading text */
+        #loading-text {
+            font-size: 24px;
+            font-weight: bold;
+        }
+    </style>
+    <script>
+        // JavaScript function to update row color
+        function updateRowColor(selectElement, row) {
+            if (selectElement.value === "") {
+                row.classList.remove("selected");
+            } else {
+                row.classList.add("selected");
+            }
+        }
+
+        // JavaScript to hide the splash screen after 3 seconds and display the main content
+        setTimeout(function () {
+            document.getElementById("splash-screen").style.display = "none";
+            document.getElementById("main-content").style.display = "block";
+        }, 2000); // 3000
+    </script>
+</head>
+
+<body>
+
+    <h1>Active Learning POC</h1>
+    <form method="POST">
+        <h3>Active Learning Step #{{ al_step }} - Remain data: {{n_unlabeled}} / 100</h3>
+        <div id="splash-screen">
+            <div id="loading-text">Active Learning Step #{{ al_step }}</br></br>{{al_process}}</div>
+        </div>
+        <div id="main-content" style="display: none;">
+            <table>
+                <colgroup>
+                    <col style="width: 80%;">
+                    <col style="width: 20%;">
+                </colgroup>
+                <tr>
+                    <th>Text</th>
+                    <th>Label</th>
+                </tr>
+                {% for item in data.text %}
+                <tr>
+                    <td>{{ item }}</td>
+                    <td>
+                        <select name="annotations" onchange="updateRowColor(this, this.parentNode.parentNode)">
+                            <option value="">None</option>
+                            {% for label in labels %}
+                            <option value="{{ label }}">{{ label }}</option>
+                            {% endfor %}
+                        </select>
+                    </td>
+                </tr>
+                {% endfor %}
+            </table>
+        </div>
+        <button type="submit">Next Samples</button>
+    </form>
+</body>
+
+</html>

templates/upload.html

@@ -0,0 +1,54 @@
+<!DOCTYPE html>
+<html>
+
+<head>
+    <title>Active Learning POC</title>
+
+    <link rel="stylesheet" type="text/css" href="{{ url_for('static', filename='css/styles.css') }}">
+
+    <style>
+        input[type=file] {
+            width: 350px;
+            max-width: 100%;
+            color: #444;
+            padding: 5px;
+            background: #fff;
+            border-radius: 10px;
+            border: 1px solid #555;
+        }
+
+        input[type=number] {
+            width: 100px;
+            max-width: 100%;
+            color: #444;
+            padding: 5px;
+            background: #fff;
+            border-radius: 10px;
+            border: 1px solid #555;
+        }
+    </style>
+</head>
+
+<body>
+    <h1>Active Learning POC</h1>
+    <form method="POST" enctype="multipart/form-data">
+        <h3>Upload Data File (CSV)</h3>
+        <input type="file" name="data_file" accept=".csv" required>
+
+        <h3>Upload Labels File (CSV)</h3>
+        <input type="file" name="labels_file" accept=".csv" required>
+        <table style="border-style: hidden">
+            <tr>
+                <td style="width: 200px;">
+                    Active Learning Sample
+                </td>
+                <td><input type="number" name="query_number" required></td>
+            </tr>
+        </table>
+
+
+        <button type="submit">Start Labeling</button>
+    </form>
+</body>
+
+</html>